Resolution

[M Most]

Why is it for a Netflix original, I can't shoot 10:1 redcode on 8k Helium or DXL, but I can shoot xavc 4k on an fs-7.

Because Netflix, like all studios, has to write a spec that is somewhat generalized, and not write a separate spec for every camera in existence and all cameras that enter the market after the spec is written. A generalized spec means that specific desires are addressed (like minimum resolution being UHD, and RAW recording being used if the camera allows it), but not getting granular enough on any single device to spell out specifics in terms of maximum allowable compression. Cameras that record RAW are expected to maintain relatively low compression so that the data can be reconstructed faithfully now and in the future. Cameras that don't are generally not considered to have the same high degree of integrity with respect to the original image, and are generally not allowed as "A" cameras on high end original productions. Documentaries, reality productions, and the like are very different than original drama and comedy, and carry less stringent requirements as a result. If you wanted to shoot a Netflix drama using an FS-7 as the "A" camera, my guess is that Netflix would not be happy about that and at the very least would encourage you to use a "better" device. Or they might simply say "no."...

 

[Eric Z]

NAB booth 2015, quote from 2006...

Also - Interlacing not spoken here!

 

[Zak Ray]

Count me as one of those resolution skeptics, but also as someone very willing to learn and open to being wrong. So please, not looking for a fight, just asking questions.

My hesitation with posts like this is they seem to boil down to "more is always better". That may be true, but I prefer to see it in practice rather than theory. How much more are you getting from 8K or 6K or 4K, when your deliverable is a 2K DCP? Can we see the extra detail really be discerned on a shot in motion? Can I see a side-by-side comparison that isn't blown up to 400%?

And to be clear, it's not just a matter of "I don't want to deal with the extra data" (though that's true as well). It's also about low-light sensitivity, which is reduced every time the pixels get smaller.

Again, I agree more is better. I just don't like talking about in generalities.

 

[David Battistella]

High resolution still frames from shot footage would be one reason.

Right now i have six images blown up to 1,5 x ,5 meters taken from boith 5k and 6k stills. The show is up right now in here inFlorence as part of the San Giovanni culture festival. Im pretty sure i would not even have this showup if i had shot the material onan HDcamera.

A second reason would be stock footage houses. The 4k clips just sell at a higher price than HD.

The shift to 4k.
Creating a 4k/2k DCP is pretty much standard now as more 4k cinemas come on board. The thing i find a bit silly is when people cite data size and cost when dealing with 4k material. Im a small, indy doc maker and every film since 2011 has been finished in 4k and thats without massive budgets. So to me when i hear the cost argument it is more based on lazyness and status quo rather than genuine reason to not finish 4k.


David

 

[Les Hillis]

...Can we see the extra detail really be discerned on a shot in motion?...It's also about low-light sensitivity, which is reduced every time the pixels get smaller...

The results of the 'extra detail' can be seen as more of a qualitative difference than a quantitative one, where the effect is less about the ability to discern crazy amounts of detail, and more about overcoming things like low-res artifacts and the side-effects of the previously required methods of correcting them.

As Graeme said, "...resolution is multi-factorial and nuanced...", which means the pro's and cons of higher and lower res image capturing are as well. It's not just about how many hairs you can count at 400% or how far in you can zoom in post.

Also, it's not always the case that smaller pixels are less light sensitive. For example, the smaller pixels on the Helium sensor are actually more light-sensitive than the larger pixels on the Dragon sensor.

Edit- I agree though, would be good to actually see some specific examples. Might have a try at it myself when I can.

 

[Graeme Nattress]

Count me as one of those resolution skeptics, but also as someone very willing to learn and open to being wrong. So please, not looking for a fight, just asking questions.

My hesitation with posts like this is they seem to boil down to "more is always better". That may be true, but I prefer to see it in practice rather than theory. How much more are you getting from 8K or 6K or 4K, when your deliverable is a 2K DCP? Can we see the extra detail really be discerned on a shot in motion? Can I see a side-by-side comparison that isn't blown up to 400%?

And to be clear, it's not just a matter of "I don't want to deal with the extra data" (though that's true as well). It's also about low-light sensitivity, which is reduced every time the pixels get smaller.

Again, I agree more is better. I just don't like talking about in generalities.

It's not really extra detail. There's only so much detail you can fit in a 2k file, but what you're seeing is the better MTF, better texture and so on. If you want to see the benefits, shoot something 8k, make a 8k, 6k, 4k, 2k image downsampled from it and view each pixel-for-pixel looking at detail / texture and feel. Obviously, it's hard to compare because you're seeing a zoomed in image more and more as you get to the 8k 1:1 but it should give you enough idea of what we're talking about.

I don't know what you mean when say "It's also about low-light sensitivity, which is reduced every time the pixels get smaller." as that's not necessarily the case. Compare Helium 8k with VV 8k and the smaller Helium pixel is more sensitive.

Graeme

 

[Zak Ray]

Thanks everyone. Allow me to offer a few counterarguments (friendly ones!) :smile5:

To be clear, I'm not proposing we shoot HD. I've been shooting 4K for years and it's great. My hesitation is with 6K, 8K, 10K... whatever RED comes out with next. Especially when most theatrical projectors are still 2K and most online content distribution is (at best) 4K for the foreseeable future.

High resolution still frames from shot footage would be one reason.

I've never really subscribed to the "stills from video" argument. Not that you can't do it, or that it's not a cool feature to have, but there's a reason people still hire still photographers. It's a different skill set.

Im a small, indy doc maker and every film since 2011 has been finished in 4k and thats without massive budgets. So to me when i hear the cost argument it is more based on lazyness and status quo rather than genuine reason to not finish 4k.

The tradeoff of an indie filmmaker has always been cost for time. True, we can afford to finish in 4K raw, but the extra time it takes to copy footage, create proxies, do final conform, and process/render the raw footage is where we make our sacrifice (because we love doing it, so it's worth the time!).

If you want to see the benefits, shoot something 8k, make a 8k, 6k, 4k, 2k image downsampled from it and view each pixel-for-pixel looking at detail / texture and feel.

Unfortunately I don't have an 8K camera to do so, but I'd love to see such a test from RED or another user. Again, I'm not suggesting there's no improvement, only wondering about how much of an improvement it really is, and balancing that with the tradeoffs.

I don't know what you mean when say "It's also about low-light sensitivity, which is reduced every time the pixels get smaller." as that's not necessarily the case. Compare Helium 8k with VV 8k and the smaller Helium pixel is more sensitive.

This is an area where I may be underinformed, but it's always been my understanding that bigger photosites = better photon collection = better sensitivity to light. Hence Sony and Canon releasing cameras these days in "low light or high megapixel" models (a7s vs a7r). I don't mean to say that one camera's pixel size necessarily determines it's sensitivity compared to another camera's pixel size, but rather that all other things being equal, larger pixels increase sensitivity.

Again though, my expertise here is minimal. Appreciate your taking the time to discuss, Graeme.

 

[Graeme Nattress]

To take your first point, just time travel back 10 years and you'll find the same argument, but it was against 4k rather than HD. Today if you want the best HD image it's being made from a much higher resolution downsampled. In computer graphics although you might render out HD, internally the math is such that it comes from a heavily oversampled internal image before being output.

Stills from motion is interesting: I'll refer you here: http://blog.vincentlaforet.com/2017...ion-converge-without-apologies-or-qualifiers/

If you want the 8k benefit with much less overhead - shoot to one of the downsampled video formats - ProRes or DnX in camera. You'll get great downsampling benefits and greatly reduce your production overhead.

Sensitivity is interesting and often counter intuitive. Just think rainfall measure. You get 3cm say of rainfall no matter how big the bucket is (assuming straight sides). There are issues with super-small cell phone pixels, but that's not what we're talking about here. What really matters is pixel design

Graeme

 

[Phil Holland]

Graeme, I think one interesting thing about resolution that hasn't come up is it's effect on tones and color. This topic is something I haven't seen explored much, but is certainly a reality when it comes to captured color detail.

 

[Zak Ray]

Yes, I saw that Vincent Laforet video. To me it's more of an argument for still photographers owning REDs rather than cinematographers shooting stills. Personally, I don't think either is going to happen in a widespread way. Perhaps I'm underestimating the number of people who shoot stills and motion on the same job.

To take your first point, just time travel back 10 years and you'll find the same argument, but it was against 4k rather than HD. Today if you want the best HD image it's being made from a much higher resolution downsampled.

That's totally fair, and I'm happy to have my 4K video downsampled from 8K. But do I really have that option? Shooting 4K on a RED can only be done by sensor cropping, meaning if I want 4K on a Helium, I have to deal with a 1.7x crop. So the only way to shoot at Super35 is to shoot twice as much data as I ultimately need, and convert it in post (extra time and money).

If you want the 8k benefit with much less overhead - shoot to one of the downsampled video formats - ProRes or DnX in camera. You'll get great downsampling benefits and greatly reduce your production overhead.

This is a good idea. Though I've read that the in-camera ProRes doesn't look as good as when it's converted via RedCineX. Is it possible to shoot just ProRes without the R3Ds?

 

[Reuben Evans]

For me it is a simple as The stuff we just shot on Helium Weapon (processed via IPP2) crushes anything we shot on Scarlet MX or Weapon MG. Resolution is just one aspect of the amazing image. I had people walk by my desk and their eyes literally popped-out! The Helium Weapon footage just looks way better—and you truly feel it. It makes you more excited, and is down right breath taking. That kind of technology I find inspirational, in the sense that I want to make more and do more because of it. It's like a challenge to step up to what the technology can do and use it to move people's hearts. Maybe that's a little superlative, but until you've shot with the camera for a little while it's hard to explain

 

[Graeme Nattress]

Graeme, I think one interesting thing about resolution that hasn't come up is it's effect on tones and color. This topic is something I haven't seen explored much, but is certainly a reality when it comes to captured color detail.

Resolution certainly has a positive benefit on tones, hues and colours. Micro-detail on hues works against homogeneity in a natural manner and the proper texture adds verisimilitude. Resolution allows for better capture of the lens MTF and thus will result in more pleasing and less artificial tones. Same with the lower propensity for aliasing.

Graeme

 

[Graeme Nattress]

Sure you can! And yes, it's perhaps not quite as good as you can do better quality downsampling in post on an R3D and there's the new highlight extension if you clip, but if you're worried about data cost and processing it's a viable option.

Graeme

 

[Graeme Nattress]

For me it is a simple as The stuff we just shot on Helium Weapon (processed via IPP2) crushes anything we shot on Scarlet MX or Weapon MG. Resolution is just one aspect of the amazing image. I had people walk by my desk and their eyes literally popped-out! The Helium Weapon footage just looks way better—and you truly feel it. It makes you more excited, and is down right breath taking. That kind of technology I find inspirational, in the sense that I want to make more and do more because of it. It's like a challenge to step up to what the technology can do and use it to move people's hearts. Maybe that's a little superlative, but until you've shot with the camera for a little while it's hard to explain

Yes, it's a tad hard to describe but we're all having our go at it! And it's possible to demonstrate some of the technical aspects, but you really have to see it all together on your own footage to really appreciate it. And yeah, the 8k leads to super-fine micro-detail which can be rather compelling when processed through a sensitive image processing pipeline designed to bring out the best in the footage.

Graeme

 

[Zak Ray]

For me it is a simple as The stuff we just shot on Helium Weapon (processed via IPP2) crushes anything we shot on Scarlet MX or Weapon MG. Resolution is just one aspect of the amazing image.

That's awesome! I totally agree there are so many things that go into a great image. I think if RED allowed for shooting different resolutions in camera without cropping, it would solve a lot of my issues.

Basically what I'm saying is, I don't see a need for 8K shooting because I don't see a need for 8K finishing (even if 8K distribution is in our near future, which I'm not convinced it is). So 8K is really only valuable to me as an oversample of 4K. But unless you want a cropped sensor, that can only be done by shooting 8K and converting in post, or shooting ProRes (which is a good option, in my opinion!)

 

[Graeme Nattress]

I think if RED allowed for shooting different resolutions in camera without cropping, it would solve a lot of my issues.

- Then it wouldn't be recording RAW....

Graeme

 

[David Battistella]

The tradeoff of an indie filmmaker has always been cost for time. True, we can afford to finish in 4K raw, but the extra time it takes to copy footage, create proxies, do final conform, and process/render the raw footage is where we make our sacrifice (because we love doing it, so it's worth the time.

Yeah, I don't know why Fincher bothered shooting "Gone Girl" in 6k.

I don't pick higher resolution because i'm an indy, I choose it because I am a professional and because there are many considerations. Like redcode raw gets better over time. I keep the first 8GB cf card of footage i shot way back on the RedOne handy. That 4k footage looks incredible now after what 9 years. Its improved. Seriously. I. And still make decent prints from those files.

Also, I'd venture a guess that the top Cinematographers on Reduser might take exception with not being considered at least very capable photographers.

David

 

[Aaron Lochert]

In computer graphics although you might render out HD, internally the math is such that it comes from a heavily oversampled internal image before being output.

In the video game industry, this is becoming a common thing, they're calling it "Resolution Scale," which is a user-defined amount of oversampling.

If that isn't a clue that oversampling is the way to go, I don't know what is.

 

[Zak Ray]

Yeah, I don't know why Fincher bothered shooting "Gone Girl" in 6k.

I don't pick higher resolution because i'm an indy, I choose it because I am a professional and because there are many considerations. Like redcode raw gets better over time. I keep the first 8GB cf card of footage i shot way back on the RedOne handy. That 4k footage looks incredible now after what 9 years. Its improved. Seriously. I. And still make decent prints from those files.

David, you may have misunderstood my post. I didn't mean to imply that 6K, 7K, 8K, and whatever comes next aren't higher quality. All I meant is that there's always a cost in shooting more data, and in the indie world, I find that cost is more often in time than in money. Hard drives may be cheap, but I think if you timed every minute that you spent in post on a raw vs ProRes production, you'd see a big difference.

Just look at all the TV shot on Alexa ProRes rather than ArriRaw... all I'm saying is there's a point of diminishing returns. So when I hear RED talking about the benefits of shooting 8K, it's not that I disbelieve them-- it's just that I want to see the benefits rather than take them at face value.

At least, that's my perspective on it. Happy to hear yours. I do appreciate your point about color science improvements over the years, that's a nice benefit of raw.

In the video game industry, this is becoming a common thing, they're calling it "Resolution Scale," which is a user-defined amount of oversampling.

If that isn't a clue that oversampling is the way to go, I don't know what is.

Again, I'm totally on board for oversampling. No issue whatsoever with that. But what's being discussed here is actually recording 6K or 8K resolution.

 

[Zak Ray]

Also, I'd venture a guess that the top Cinematographers on Reduser might take exception with not being considered at least very capable photographers.

That may be true, and I don't mean to insult the skills of any DP+photographers. I dabble in photography as well. But I think the odds of a DP getting hired to do both on a traditional production are quite small, almost as small as the odds of a pro photographer trading in his $2,000 DSLR for a $50,000 RED.